STATISTICS 538 (2004-2005, Term 1)

Course Outline (ps file)

Assigned Coursework (ps file)

R code used for examples

Lecture 6 (Sept. 27): Simulation study for probit versus logit models

Lecture 8(a) (Oct. 4): Deviance test for goodness-of-fit in Poisson regression.

Lecture 8(b) (Oct. 4): Likelihood ratio versus Wald test in Poisson regression setting.

Lecture 9 (Oct. 6): Uncollapsed versus collapsed binary data.

Lecture 10 (Oct. 13): Residuals with binomial data.

Lecture 14 (Oct. 27): Understanding proportional odds and proportional hazards models.

Lecture 15 (Oct. 27): Poisson versus multinomial modelling of count data.

Lecture 18 (Nov. 15): Quasi-likelihood (also rats data).

Lecture 19 (Nov. 17): GLMM via penalized quasi-likelihood.

Lecture 20 (Nov. 22): GLMM via Bayes/MCMC.

Lecture 21 (Nov. 24): Marginal models / GEE.

Lecture 22 (Nov. 29): EM for missing data.